Date: Wed, 9 Oct 2002 11:18:20 -0400 (EDT) From: Andrew Gallatin <gallatin@cs.duke.edu> To: Jeff Roberson <jroberson@chesapeake.net> Cc: gibbs@freebsd.org, current@freebsd.org Subject: Re: alpha tinderbox failure Message-ID: <15780.18620.457682.641298@grasshopper.cs.duke.edu> In-Reply-To: <20021008235836.B44108-100000@mail.chesapeake.net> References: <200210082312.g98NCS5R025098@beast.freebsd.org> <20021008235836.B44108-100000@mail.chesapeake.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Jeff Roberson writes:
>
>
> On Tue, 8 Oct 2002, Dag-Erling Smorgrav wrote:
> > "Makefile", line 4194: warning: duplicate script for target "geom_bsd.o" ignored
> > cc1: warnings being treated as errors
> > /h/des/src/sys/dev/advansys/adv_pci.c: In function `adv_pci_attach':
> > /h/des/src/sys/dev/advansys/adv_pci.c:197: warning: overflow in implicit constant conversion
> > *** Error code 1
>
> Any progress on this?
This particular message is caused by alpha's
#define BUS_SPACE_UNRESTRICTED (~0UL)
Clashing with int nsegments:
/* XXX Should probably allow specification of alignment */
int bus_dma_tag_create(bus_dma_tag_t parent, bus_size_t alignemnt,
bus_size_t boundary, bus_addr_t lowaddr,
bus_addr_t highaddr, bus_dma_filter_t *filtfunc,
void *filtfuncarg, bus_size_t maxsize, int nsegments,
bus_size_t maxsegsz, int flags, bus_dma_tag_t *dmat);
Sparc64 has the same problem. ia64 gets around it by just making
BUS_SPACE_UNRESTRICTED an int:
#define BUS_SPACE_UNRESTRICTED (~0)
I'd like to do the same for alpha. I think this is valid, as
BUS_SPACE_UNRESTRICTED seems to be used exlusively as an argument
to bus_dma_tag_create(... nsegments = BUS_SPACE_UNRESTRICTED...)
I'd also like to add a bus_space_subregion(). Please review the
appended patch. I'm running it with no ill effects, and it makes
alpha get a bit further on in LINT. (until it dies on printf format
warnings). Its going to be a bear to get a clean lint with Werror.
All those crusty old isa drivers casting pointers to integers make
me feel a bit overwhelmed.
It would be nice if des could remove LINT from the tinderbox builds
until LINT has a chance of compiling. Right now people are ignoring
the alpha tinderbox failure messages. Eventually, it will fail for a
reason besides useless LINT garbage, and everybody will ignore it.
Drew
Index: include/bus.h
===================================================================
RCS file: /home/ncvs/src/sys/alpha/include/bus.h,v
retrieving revision 1.11
diff -u -r1.11 bus.h
--- include/bus.h 4 Oct 2002 20:40:39 -0000 1.11
+++ include/bus.h 9 Oct 2002 00:06:39 -0000
@@ -88,7 +88,27 @@
/* The largest address space known so far is 40 bits */
#define BUS_SPACE_MAXADDR 0xFFFFFFFFFUL
-#define BUS_SPACE_UNRESTRICTED (~0UL)
+#define BUS_SPACE_UNRESTRICTED (~0)
+
+/*
+ * Get a new handle for a subregion of an already-mapped area of bus space.
+ */
+
+static __inline int bus_space_subregion(bus_space_tag_t t,
+ bus_space_handle_t bsh,
+ bus_size_t offset, bus_size_t size,
+ bus_space_handle_t *nbshp);
+
+static __inline int
+bus_space_subregion(bus_space_tag_t t __unused, bus_space_handle_t bsh,
+ bus_size_t offset, bus_size_t size __unused,
+ bus_space_handle_t *nbshp)
+{
+
+ *nbshp = bsh + offset;
+ return (0);
+}
+
struct alpha_busspace;
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?15780.18620.457682.641298>